Low Frequency of p.S510G in PIAS1 Challenges its Relevance for Modifying Repeat Expansion Disorders

We read with great interest the recent study by Lee and colleagues entitled “A PIAS1 Protective Variant S510G Delays polyQ Disease Onset by Modifying Protein Homeostasis,” reporting the p.S510G variant in the PIAS1 protein as a novel modifier of age at onset (AAO) in Huntington’s disease (HD) and spinocerebellar ataxia type 3 (SCA3). By targeted sequencing of 583 genes implicated in proteinopathies in relatively small groups of HD and SCA3 patients, the authors identified variants in 16 genes to be associated with late AAO. One of those genes was PIAS1, encoding an E3 SUMO (small ubiquitin-like modifier) ligase, that has been previously associated with HD. Therefore, the authors performed follow-up functional analyses of one of the candidate modifier variants (rs755539001, c.1528A>G, p.S510G) in PIAS1 that suggested a protective function based on decreased SUMOylation and less accumulation of the mutated Huntingtin protein containing an expanded polyglutamine tract in in vitro and in HD mouse models. Indeed, SUMO modifications have long been implicated in neurological repeat expansion and other neurodegenerative disorders. More recently, a link has been reported between PIAS1 and DNA damage repair mechanisms. This is of particular interest given that variants in several DNA repair genes have been associated with AAO in repeat expansion disorders. A recently discovered example in this context is X-linked dystonia-parkinsonism (XDP), an X-linked recessive adultonset movement disorder caused by a founder retrotransposon insertion in the TAF1 gene. The length of a polymorphic hexanucleotide repeat within this insertion accounts for 50% of the AAO variability among patients. Furthermore, two of the genes coding for DNA repair proteins, MSH3 and PMS2, contain additional genetic modifiers of AAO in XDP. Finally, XDP and HD share neuropathological similarities, because they both primarily affect the striatum and display a severe loss of medium spiny neurons. Considering all of the above, we hypothesized that the p.S510G PIAS1 variant investigated by Lee and colleagues might be a candidate protective modifier of AAO in XDP. Of note, according to the Genome Aggregation Database (gnomAD), the p.S510G variant is extremely infrequent, and it has been detected only once in over 120,000 individuals in the entire database (allele frequency: 0.000004017) and in none of 9,000 investigated East Asian individuals (https://gnomad.broadinstitute.org/ variant/15-68473596-A-G). Although Lee et al do not report the frequency of this variant in their investigated samples, it must have been comparably low. Despite its documented rarity, we investigated the presence of the p.S510G variant in a cohort of 190 XDP patients, given that population-specific frequencies could differ and the Philippines is located in relative proximity to Taiwan. However, none of our patients carried this change. Therefore, although molecular analyses of this variant yielded some promising functional data, the p.S510G variant cannot be a relevant modifier of HD, SCA3, XDP, or other repeat expansion disorders in the general population, merely because of its extreme rarity.